Michigan Law Review Michigan Law Review

Volume 105 Issue 6

2007

The Theory and Practice of Tax Reform The Theory and Practice of Tax Reform

Lawrence Zelenak Duke Law School

Follow this and additional works at: https://repository.law.umich.edu/mlr

Part of the Taxation-Federal Commons, and the Tax Law Commons

Recommended Citation Recommended Citation Lawrence Zelenak, The Theory and Practice of Tax Reform, 105 MICH. L. REV. 1133 (2007). Available at: https://repository.law.umich.edu/mlr/vol105/iss6/7

This Review is brought to you for free and open access by the Michigan Law Review at University of Michigan Law School Scholarship Repository. It has been accepted for inclusion in Michigan Law Review by an authorized editor of University of Michigan Law School Scholarship Repository. For more information, please contact mlaw.repository@umich.edu.

THE THEORY AND PRACTICE OF TAX REFORM

Lawrence Zelenak*

SIMPLE, FAIR, AND PRO-GROWTH: PROPOSALS TO Fix AMERICA'S TAX

SYSTEM. By The President's Advisory Panel on Federal Tax Reform. 2005.Pp. xv, 272. Available at http://www.taxreformpanel.gov/final-report/.

INTRODUCTION

On January 7, 2005, President Bush-flush with recent electoral vic-tory-issued an Executive Order creating the President's Advisory Panel onFederal Tax Reform. The Order instructed the bipartisan Panel to recom-mend one or more plans for major reform of the federal income tax . Thepresident did not, however, permit the Panel to begin its work on a blankslate. Instead, the Order required (among other things) that the Panel's pro-posals be revenue-neutral, simpler than current law, "appropriatelyprogressive," and supportive of homeownership and charity Although theOrder contemplated that the Panel might offer more than one tax reformplan, it required that at least one option "use the Federal income tax as thebase for its recommended reforms. 4

The Panel5 went to work promptly, held several hearings around thecountry, and produced a Report (not quite on schedule 6) in November 2005.With Congress having difficulty enacting any legislation and with its limitedlegislative capacity focused on issues more pressing than revenue-neutraltax reform, the Report was widely viewed as dead on arrival.7 The Report is,however, a substantial and thoughtful piece of work, and when Congress isfinally able to turn its attention to tax reform sometime in the next fewyears, it can and should look to the Report for ideas---especially since theReport's recommendations were largely driven by the Panel's determinationto repeal the alternative minimum tax (AMT), and repeal (or at least major

* Pamela B. Gann Professor of Law, Duke Law School.

I. Exec. Order No. 13,369, 70 Fed. Reg. 2323 (Jan. 7, 2005).

2. Id.

3. Id.

4. Id.

5. The Co-Chairs of the Panel were former senators Connie Mack and John Breaux. Theother Panel Members were William E. Frenzel, Elizabeth Garrett, Edward P. Lazear, Timothy J.Muris, James M. Poterba, Charles 0. Rossotti, and Liz Ann Sonders. P. iii.

6. The original Order had called for the submission of the report no later than July 31,2005. Exec. Order No. 13,369, 70 Fed. Reg. at 2324.

7. Steven Pearlstein, Tax Reform That's Bold and Beautiful, WASH. POST, Nov. 4, 2005, atDI.

1133

Michigan Law Review

reform) of the AMT is almost certain to be an issue of great legislative inter-est in the not-too-distant future. 8

Part I of this Review outlines the basics of the two proposals that thePanel presents in the Report. Part II evaluates some of the proposals' details.Finally, Part III concludes with two comments on the place of the Panel'sReport within the long-standing tax reform debate.

I. THE PANEL'S REPORT

Following the Executive Order's suggestion that the Panel might offermore than one reform plan, the Panel's Report endorses both a "SimplifiedIncome Tax Plan" (SITP) and a "Growth and Investment Tax Plan" (GITP).Although the two plans have slightly different rate structures-brackets of15%, 25%, 30%, and 33% under the SITP, and brackets of 15%, 25%, and30% under the GITP-and differ significantly in their treatment of businessand investment income, they have many features in common (p. 61 tbl.5.1).Each would repeal the alternative minimum tax (pp. 85-87); replace thestandard deduction and personal exemptions with a "family credit," therebyeliminating the distinction between "above-the-line" deductions available toall taxpayers and itemized deductions available only to taxpayers not claim-ing the standard deduction (pp. 63-68); greatly reduce marriage penalties bygiving married taxpayers a tax rate schedule with brackets twice as wide asthose applicable to single taxpayers and by giving married taxpayers a "fam-ily credit" twice as large as a single person's credit (p. 89); eliminate thededuction for state and local taxes (pp. 83-84); replace the home mortgageinterest deduction with a 15% credit, while capping the loan principalamount generating credit-eligible interest at the average cost of housingwithin the taxpayer's area (resulting in limits ranging from about $227,000to $412,000) (pp. 70-75); make the charitable contribution deduction avail-able to all taxpayers, but only to the extent contributions exceed 1% ofincome (pp. 75-76); cap the exclusion for employer-provided health insur-ance at the average cost of health insurance, while allowing a similarlycapped deduction for the purchase of health insurance outside the employ-ment context (pp. 78-82); and greatly expand opportunities for tax-favoredsavings outside the employment context by allowing every taxpayer to con-tribute up to $10,000 per year to a "Save for Retirement" (SFR) account andanother $10,000 per year to a "Save for Family" (SFF) account, both basedon the "tax pre-paid" Roth IRA model.9 Both proposals retain tax-favored

8. See infra Section H.C. For the gory details on the impending AMT train wreck in theabsence of legislative intervention, see STAFF OF J. COMM. ON TAX'N, 109TH CONG., PRESENT LAWAND BACKGROUND RELATING TO THE INDIVIDUAL ALTERNATIVE MINIMUM TAX (Comm. Print2005).

9. Pp. 119-21. 159. Under the Roth IRA model, contributions to tax-favored accounts arenot deductible and distributions from the accounts are not taxable. Eligibility to make contributionsto these "tax pre-paid" accounts would not be phased out for high-income taxpayers and would notbe denied to taxpayers participating in tax-favored employer-sponsored savings plans. Pp. 119-20.

1134 [Vol. 105:1133

The Theory and Practice of Tax Reform

treatment for employer-sponsored retirement savings, but the tax favoritismin the SITP takes the familiar form of cash flow treatment (under which notax is imposed at the time contributions are made, but tax is imposed whenthe taxpayer receives account distributions), while the GITP applies the "taxpre-paid" Roth IRA model to employer-sponsored plans (pp. 115-19, 159-60).

The biggest differences between the two plans are in their treatment ofbusiness and investment income. Although the SITP has very significantconsumption tax features, including the tax-favored savings vehicles de-scribed above, it retains the norm of depreciation, rather than immediateexpensing, for most long-lived business assets.0 The GITP, in sharp con-trast, adopts the consumption tax approach of immediate expensing for allbusiness assets (pp. 163-64). Although the Panel's Report does not claimthe GITP is an income tax, it is also not a pure consumption tax because itretains a modest amount of taxation of investment income-a 15% rate ap-plicable to dividends, capital gains, and interest (pp. 152, 159).

II. LOOKING AT SOME DETAILS

A. Tax Favored Savings, Wage Taxes, Cash Flow Taxesand the Budget Window

Despite its retention of the "income tax" label, the SITP would consti-tute a significant move in the consumption tax direction-primarily as aresult of the introduction of SFR and SFF accounts. Under a consumptiontax, income saved in one year is not taxed until it is spent on consumption insome later year. The current "income" tax follows this consumption tax-or"cash flow"-model with respect to employer-sponsored retirement savingsand traditional individual retirement accounts (IRAs)." If a taxpayer facesthe same tax rate in both the year in which income is saved and the subse-quent year in which the income (augmented by the investment return on thesavings) is consumed, the results to the taxpayer will be the same under ei-ther a cash flow model (with tax imposed only in the year of consumption)or a wage tax model (with tax imposed only in the year of saving). 12 Unlike

Withdrawals from "Save for Family" accounts could be made to pay for retirement, health, educa-

tion, or the down payment on a home. P. 120.

10. Pp. 131-32. The SITP would eliminate most double taxation of corporate income byproviding an exclusion for dividends paid out of corporate profits subject to U.S. tax and an exclu-sion for 75% of the capital gain on the sale of stock of a U.S. corporation. Pp. 124-25.

11. See I.R.C. § 402(a) (2000) (employer-sponsored plans); id. § 408 (IRAs). Current law

does not precisely follow the consumption tax model, in that it imposes tax on distributions fromemployer-sponsored retirement plans and IRAs whether or not the taxpayer spends the distributedamounts in the year of distribution. However, to the extent retirees do spend distributed amounts oncurrent consumption, the tax result is consistent with a consumption tax.

12. For example, suppose that a taxpayer subject to a 20% tax rate at all relevant times wantsto devote $10,000 of this year's earnings to retirement savings, and that any amounts saved this yearwill triple in value by the retirement year in which the taxpayer consumes the savings. Under thecashflow approach, he will be able to save the entire $10,000 (because no current tax is owed on

1135April 20071

1136 Michigan Law Review [Vol. 105:1133

employer-sponsored retirement plans and traditional IRAs, Roth IRAs aretaxed under the wage tax model-but Roth IRAs constitute only a smallportion of tax-favored retirement savings. 4

The proposed SFR and SFF accounts both follow the wage tax modelrather than the dominant cash flow model. Although they depart from theusual technique for introducing non-income tax features into the "incometax," the accounts would constitute a major increase in the hybridization ofthe so-called income tax. 5 If a married couple were to save the maximumamount of $40,000 per year in the two new accounts ($10,000 for eachspouse in each type of account), and the invested funds grew at the rate of5% per annum, at the end of thirty years they would have over $2.6 millionin their accounts, which they could consume free of tax. The SITP wouldretain the current ceiling on annual contributions to employer-sponsoredretirement plans (which would continue to be taxed under the cash flowmodel) (p. 117). If each spouse saved the annual maximum of $44,000 atwork 6 each year for thirty years, and the rate of return on the investmentwas again 5%, at the end of thirty years they would have over $5.7 millionin their employer-sponsored plans (which would, however, be taxed upondistribution). When wealth of this magnitude is afforded wage tax or cashflow treatment, the continued application of the "income tax" label is dubi-ous.

As interesting as the Panel's decision to make the SITP less of an in-come tax is its decision to use wage tax treatment rather than cash flow

saved income), which will grow to $30,000 by the year of consumption. In that year he will pay$6,000 tax (20% of $30,000), and he will be able to spend the remaining $24,000 on consumption.Under the wage tax alternative, he will be able to save only $8,000 (because he must devote theother $2,000 to the tax on the $10,000 of earnings), but that will grow to $24,000 by the year ofconsumption, and he will be able to spend the entire $24,000 on consumption in that year. Where t,represents the tax rate in the year of savings and t2 represents the tax rate in the year of consumption,and any amounts invested will triple in value between the two years, the formula for the amount thetaxpayer can consume in the later year under cash flow taxation is $10,000 x 3 x (1-t 2), and theformula for the amount the taxpayer can consume in the later year under wage taxation is $10,000 x(1 - t) x 3. As long as t, and t2 are identical, the two formulas produce the same number.

13. I.R.C. § 408A; see also id. § 402(c) (permitting taxpayers to elect wage tax treatment, inlieu of cash flow treatment, for elective contributions to employer-sponsored plans).

14. This is true even when the focus is limited to savings outside the employment context. In2005, 37.6 million households owned traditional IRAs, while only 16.1 million households ownedRoth IRAs. Sandra West & Victoria Leonard-Chambers, The Role of IRAs in Americans'RetirementPreparedness, RESEARCH FUNDAMENTALS (Inv. Co. Inst., Wash. D.C.), Jan. 2006, at 1, 2 fig.2.

15. The proposed accounts would replace a number of existing specialized vehicles for tax-favored retirement savings, including IRAs, Roth IRAs, health savings accounts, Coverdell savingsaccounts (for higher education), and qualified tuition programs. Pp. 119-21. Given the restrictiveeligibility requirements and contribution limits for these existing programs, the net effect of repeal-ing the existing programs and replacing them with the two proposed accounts would be a largeincrease in the availability of tax-favored savings vehicles.

16. I.R.C. § 415(c); I.R.S. Notice 2005-75, 2005-45 I.R.B. 929 (inflation-adjusted contribu-tion ceiling for 2006).

The Theory and Practice of Tax Reform1

treatment to accomplish that goal. The Panel's other proposal, the GITP,exhibits even greater enthusiasm for wage tax treatment; it would replacecash flow taxation of employer-sponsored savings with wage taxation (pp.159-60). At first glance, this preference for wage taxes over cash flow taxesis difficult to understand. After all, cash flow taxation has been the norm fortax-favored savings under the federal income tax for many decades, withapparently satisfactory results. Given the fact that the two approaches gener-ally produce similar or identical results in the end, what could explain thePanel's preference? One reason might be cosmetics. For a taxpayer in the33% bracket (the top bracket under the SITP), $10,000 in a wage tax ac-count is as good as $14,925 in a cash flow account.17 If the Panel feared thatcash flow account contribution ceilings of almost $15,000 would appear toogenerous, it might have opted for wage tax treatment to reduce the nominalcontribution ceilings to $10,000 without any change in actual generosity.

One would hope, however, that the Panel had a better reason than thatfor its rejection of the long-standing preference for cash flow taxation. Thereis a substantive difference between wage taxes and cash flow taxes when-ever a taxpayer is not subject to the same tax rate in the year income isearned and saved, and the later year in which the savings are consumed.Under a wage tax, the tax liability is a function of the tax rate in the earlieryear, while under a cash flow tax it is a function of the tax rate in the lateryear. The Panel might have believed, for some reason, that the tax rate in theearlier year is the more appropriate rate to apply when the two rates differ.In fact, however, the Report expresses no such belief.'8 It provides no expla-nation whatsoever for the choice of wage taxation rather than cash flowtaxation for the SFR and SFF accounts.' 9

The Report is slightly more forthcoming with respect to the proposal, un-der the GITP, to switch to a wage tax model for employer-sponsored savings.The Report claims to favor the wage tax model "on policy grounds," but theonly policy it actually mentions is that wage taxation "has the advantage ofbeing simpler because the traditional IRA approach involves claiming a de-duction when money is contributed and reporting income when the money iswithdrawn" (p. 160). Even if the factual premise were correct, this simplic-ity argument would be very weak (because claiming IRA deductions is notcomplicated). But, in fact, the premise is wrong. The up-front tax benefit for

17. This is because the tax has already been paid in the case of the wage tax account (a tax of$4925, which is 33% of pre-tax wages of $14,925), whereas the tax has yet to be paid in the case ofthe cash flow account. If the taxpayer withdraws the entire $10,000 from the wage tax account, thetaxpayer will owe no tax and can spend the entire $10,000. If the taxpayer withdraws the entire$14,925 from the cash flow account, the taxpayer will owe tax of $4925 and will be able to spendonly the remaining $10,000.

18. This is not surprising in the case of the SITP, since the SITP would retain cash flowtreatment for employer-sponsored retirement savings. If the Panel had asserted the superiority ofwage taxation for savings outside the employment context, it would have had to explain why thesame logic did not apply within the employment context.

19. Pp. 119-21 (describing the proposal without explaining the rationale for the wage taxchoice).

1137April 2007]

Michigan Law Review

employer-sponsored savings under a cash flow tax is an exclusion, not adeduction, and from the employee's point of view nothing could be simplerthan an exclusion.

The real reason for the Panel's preference for wage taxation over cashflow taxation is mentioned in the Report-as a piece of information, butwithout any acknowledgment that it is the reason for the preference. TheReport explains that the Panel interpreted the president's revenue-neutralitydirective to require only that the proposals be revenue-neutral over a "budgetwindow" consisting of the first ten years a proposal is in effect (p. 46). TheReport also explains that the SFR and SFF accounts would produce only aslight reduction in revenue within the ten-year window; taxpayers contribut-ing to the accounts would pay tax on wages when earned, and sometaxpayers would elect to make taxable conversions of existing cash flowaccounts to wage tax accounts (p. 47). Outside the budget window, however,the revenue loss would grow rapidly, as taxpayers received tax-free distribu-tions from wage tax accounts instead of taxable distributions from cash flowaccounts. Citing the work of Treasury Department revenue estimators, theReport explains: "[F]or [wage tax] retirement accounts, the revenue costduring the ten-year budget window is roughly one-third of the total revenuecost of this program; two-thirds of the revenue loss is not reflected in therevenue tables provided in this report" (p. 47). Although the Report care-fully avoids offering the budget window as a reason for preferring wage taxtreatment over cash flow taxation, 20 it is obvious-almost embarrassinglyobvious-that this is the one and only reason for the Panel's embrace ofwage taxation. It is deeply disappointing that a bipartisan Panel, operatingfree of political constraints, would resort to budget flim-flammery to makeits proposals appear revenue-neutral when they really are not, abandoning adecades-long preference for cash flow taxation in the process. On the otherhand, perhaps the Panel deserves some credit for being so disarmingly hon-est about its dishonesty. Without ever admitting that its preference for wagetaxation is solely a matter of budget gimmickry and that its proposals are,therefore, not really revenue-neutral, the Report provides a thorough andlucid description of the gimmickry it employs. It is a strange performance,perhaps reflective of a compromise between Panelists who did not want touse budget gimmickry at all and those who wanted to use gimmickry with-out calling attention to it.

20. In its discussion of wage tax treatment of employer-sponsored retirement savings underthe GITP, the Report mentions the budget window difference between wage taxation and cash flowtaxation, but it does not offer that difference as a reason for favoring wage tax treatment. P. 160.Similarly, in a discussion of the SFR and SFF accounts, the Report mentions that wage tax treatmentand cash flow treatment "have different near-term tax revenue consequences," but it does not sug-gest that difference is a reason to prefer one approach over the other. P. 93.

1138 [Vol. 105:1133

The Theory and Practice of Tax Reform

B. Marriage Penalty Relief

The Report recommends reducing marriage penalties in the tax-ratestructure by giving married couples who file joint returns rate brackets twiceas wide as the rate brackets for single taxpayers.2' Similarly, the familycredit (the replacement for personal exemptions and the standard deduction)would be twice as large for a married couple as for a single person (p. 89).The Report offers exactly two sentences in support of these proposals:"Couples find it hard to understand why they should pay more in tax afterthey get married than they would have paid if they had remained single,"and "By providing marriage penalty relief, the Panel's options help reducethe barriers faced by potential second earners" (p. 89).

The first sentence is fine as far as it goes, but it does not go nearly farenough. Unmentioned by the Report, the federal income tax has a long andnot very happy history of disputes over the relative treatment of married and

22unmarried taxpayers. The approach proposed by the Report has been triedbefore: it was introduced into the income tax in 1948.23 The 1948 legislationproduced no marriage penalties, but many marriage bonuses. Every mar-riage bonus can be seen, from the other side of the coin, as a singles penalty.Single persons complained, and in 1969 they were somewhat mollified bylegislation making the rate brackets for single persons more than half the

24width of the joint-retum brackets. The result was a system with both mar-riage penalties and marriage bonuses, and the basic compromise of 1969 hascontinued (albeit with considerable tweaking) to this day. It is not necessarilyunreasonable to propose a return to the 1948 approach, but it is unreason-able to propose it so cavalierly, without any discussion of the difficulty ofthe problem and the objections single persons will certainly raise to the pro-posed solution.

As with the issue of wage taxation versus cash flow taxation, the Reporthelpfully provides the information necessary to understand the inadequacyof its marriage penalty policy analysis. The Report includes a table with anumber of examples of how the SITP would affect the tax liabilities ofhypothetical taxpayers (p. 142 tbl.6.6), showing that married taxpayerswould fare much better than both single taxpayers and heads of households.For example, for taxpayers (under age 65) at the ninety-fifth percentile ofthe income distribution, the SITP would result in a decrease of 8.1% in thetax liability of a married couple, an increase of 5.9% in the tax liability of asingle person, and an increase of 15.2% in the tax liability of a head of

21. P. 89. This would eliminate rate bracket marriage penalties for married couples vis-A-vissingle taxpayers other than heads of households.

22. For a brief review of the history, see Lawrence Zelenak, Doing Something About Mar-riage Penalties: A Guide for the Perplexed, 54 TAX L. REV. 1, 4-8 (2000).

23. Revenue Act of 1948, Pub. L. No. 80-471, §§ 301-305, 62 Stat. 110, 114-16.

24. Tax Reform Act of 1969, Pub. L. No. 91-172, § 803, 83 Stat. 487, 678-85.

April 2007] 1139

Michigan Law Review

household. 2' The Report proposes a massive shifting of the income tax

burden from married taxpayers to unmarried taxpayers, with virtually nodiscussion of the merits of the proposal.

The second sentence of the Panel's justification for its proposal-about"reduc[ing] the barriers faced by potential second eamers"-seems to bebased on a misconception. Joint filing by married couples raises two differ-ent concerns. First, when two people get married, their tax liability maychange as a result of the change in their filing status, even though the in-come of each person remains the same. When marriage causes an increasein combined tax liability, that is a marriage penalty. The proposal wouldindeed decrease marriage penalties, albeit (as described above) at a consid-erable cost to unmarried persons. The second concern is the "stackingeffect." When two people are already married and one spouse (typically thehusband) is firmly committed to the workforce, while the other spouse(typically the wife) is deciding between paid work and full-time homemak-

26ing, joint returns will discourage the latter spouse from employment. If shetakes a job, her income will be "stacked" on top of her spouse's existingincome, and all her income will be subject to a relatively high marginal taxrate. 7 Separate filing by married couples would eliminate the stacking ef-fect, and a two-earner deduction would reduce it,2s but making joint-returnrate brackets twice as wide as the brackets applicable to single taxpayers hasno impact on it. 9

C. The Alternative Minimum Tax and the Deductionfor State and Local Taxes

The alternative minimum tax (AMT) functions as a shadow income tax,with a broader base than the regular income tax and a different rate structure

25. P. 142 tbl.6.6. For taxpayers at the fiftieth percentile of the income distribution, a singleperson would receive a tax reduction of 4.0%, a head of household would receive a minuscule re-duction of 0.4%, and a married couple would receive a reduction of 30.9%. P. 142 tbl.6.6.

26. See Lawrence Zelenak, Marriage and the Income Tax, 67 S. CAL. L. REv. 339, 365-72(1994) (describing this phenomenon and labeling it the "stacking effect"); see also Zelenak, supranote 22, at 19-21, 34 (noting the common confusion between marriage penalties and the stackingeffect, and explaining that the 1948 income tax system-with joint return rate brackets twice aswide as the rate brackets for unmarried taxpayers--produced the stacking effect despite producingno marriage penalties).

27. Of course, nothing in the Internal Revenue Code specifies that one spouse's income getsthe benefit of the lower rates while the other spouse's income is stacked on top and subject to higherrates. However, if one spouse is firmly committed to the workforce and the other is at the workforce-homemaking margin, the spouses will act as if the Code so specified.

28. See Economic Recovery Tax Act of 1981 (former I.R.C. § 221), Pub. L. No. 97-34,§ 103, 95 Stat. 172, 187-88 (repealed 1986).

29. Increasing the width of the joint-return's lower rate brackets would decrease the marginaltax rates faced by some spouses considering entry into the work force, thus decreasing the stackingeffect for those spouses. The stacking effect has nothing to do, however, with the width of joint-return rate brackets relative to the width of the rate brackets for unmarried taxpayers, and so hasnothing to do with marriage penalties per se.

1140 [Vol. 105:1133

The Theory and Practice of Tax Reform

(featuring a higher exemption and a lower top rate). ° If a taxpayer's tenta-tive minimum tax under the AMT rules exceeds the taxpayer's regular taxliability, he must pay both his regular tax liability and the excess of the ten-tative minimum tax over the regular tax." When the predecessor of thecurrent AMT was enacted in 1969,32 it was designed to affect only a smallnumber of rich taxpayers who made extensive use of tax preferences, andthe tax worked as intended for decades.33

The AMT, however, is now becoming a monster, with its claws reachingdeep into the middle class. The Panel wrote its Report in the shadow of aprediction by the Staff of the Joint Committee on Taxation that in 2006 theAMT would affect 18.4% of taxpayers in the $75,000-$100,000 incomerange, and 62.6% of taxpayers with incomes in the $100,000-$200,000range.34 The growth in the AMT is largely explained by three factors: (1) theexemption amounts and the rate structure of the regular tax are indexed forinflation, but those of the AMT are not; (2) the large reductions in the regulartax enacted during the Bush administration have not been matched by corre-sponding cuts in the AMT;5 and (3) the AMT disallows several decidedlymiddle-class tax benefits, including dependency exemptions and the deduc-tion for state and local taxes. The AMT monster will continue to grow in theabsence of legislative action; by 2010 the AMT is likely to apply to morethan one-third of taxpayers with incomes from $50,000 to $75,000, and al-most three-quarters of taxpayers with incomes from $75,000 to $100,000.36

The Panel's Report notes that the AMT is expected to produce more than$1.2 trillion of tax revenue over the next ten years (p. 43), and that by 2013the AMT without the regular tax would produce more revenue than the regu-lar tax without the AMT (p. 87).

30. I.R.C. §§ 55-58 (2000 & Supp. 2006).

31. The effect is the same as requiring the taxpayer to pay the greater of the regular tax orthe tentative minimum tax.

32. Tax Reform Act of 1969, Pub. L. No. 91-172, § 301, 83 Stat. 487, 580-86.

33. See, e.g., S. REP. No. 99-313, at 518 (1986) (describing the alternative minimum tax asbeing designed "to ensure that no taxpayer with substantial economic income can avoid significanttax liability by using exclusions, deductions, and credits").

34. STAFF OF J. COMM. ON TAX'N, 109TH CONG., PRESENT LAW AND BACKGROUND RELAT-

ING TO THE INDIVIDUAL ALTERNATIVE MINIMUM TAX 13 tbl.3 (Comm. Print 2005). Legislationenacted after the Panel issued its Report provided significant AMT relief for 2006, but that legisla-tion merely delays the AMT train wreck by one year. Tax Increase Prevention and ReconciliationAct of 2005, Pub. L. No. 109-222, § 301, 120 Stat. 345, 353 (2006). (Despite the name of the Act, itdid not become law until 2006.)

35. The two most significant pieces of Bush-era tax cut legislation are the Economic Growthand Tax Relief Reconciliation Act of 2001, Pub. L. No. 107-16, 115 Stat. 38 (codified as amended inscattered sections of 26 U.S.C.); and the Jobs and Growth Tax Relief Act of 2003, Pub. L. No. 108-27, 117 Stat. 752 (codified as amended in scattered sections of 26 U.S.C.).

36. Leonard E. Burman et al., The AMT. Projections and Problems, 100 TAx NOTES 105, 110tbl.2 (2003) (projecting an AMT impact in 2010 on 36.6% of taxpayers in the $50,000-$75,000income range, and on 72.9% of taxpayers in the $75,000-$100,000 income range).

1141April 2007l

Michigan Law Review

The Report brands the AMT as "the most vivid example of the wastefulcomplexity that has been built into our system," and forcefully recommendsits repeal (pp. 85-86). The Report's condemnation of the AMT is based noton the substantive design of the AMT, viewed as a free-standing tax system,but rather on the complexity of subjecting taxpayers simultaneously to twoparallel systems: "Eliminating the AMT would free millions of middle-classtaxpayers-21.6 million in 2006 and 52 million in 2015-from filing theforms, preparing the worksheets, and making the seemingly endless calcula-tions required to determine their AMT liability.' 37

If the only reason for rejecting the use of parallel tax systems is returnpreparation complexity, perhaps AMT repeal is not so urgent after all. In taxyear 2003, paid preparers were responsible for 62% of all individual taxreturns, with another 25% of returns being prepared by taxpayers with theuse of tax preparation software; only 13% of returns were prepared by tax-payers themselves without the use of software. 0 Joseph Thorndike hassuggested that this makes the return preparation complexity allegedlycaused by the AMT a non-issue: "The AMT is shaping up to be the nation'snext great nondisaster. Move over, Y2K (the global computer nonevent of2000)-the AMT wants a spot in the Chicken Little Hall of Fame."3 9 If the"seemingly endless computations," which so concerned the Panel, are han-dled by H&R Block or by TurboTax, they are not a problem, and the nationcan continue with its parallel taxes indefinitely. Thorndike predicts the AMTwill not "cause enough pain to ensure its own demise. 40 He is not happyabout this because he views the AMT as "a-bad tax, a blight on the nation'srevenue structure."4' Oddly, however, he does not explain what is wrongwith having parallel taxes if they do not result in computational headaches.

The question is significant because the Panel urges repeal of the AMTbased on its tax return preparation burden, and it appears the Panel hasgreatly exaggerated that burden. My sympathies are with the Panel, how-

37. P. 86. Of course, if the objection is only to the complexity of requiring taxpayers to dealwith two parallel systems, it might make just as much sense to repeal the regular tax and to keep areformed version of the AMT, rather than repealing the AMT and keeping a reformed version of theregular tax. The Report considers this possibility, but rejects it because of what the Panel views assubstantive design flaws in the AMT-including the absence of inflation indexing, large marriagepenalties, and the disallowance of personal and dependency exemptions. P. 87. However, the Panel'sSITP proposal could be viewed as a reformed version of a stand-alone AMT rather than a reformedversion of the regular tax, given that it shares with the AMT a broader base and a lower top marginalrate than the regular tax-especially since both the SITP and the AMT (but not the regular tax)allow no deduction for state and local taxes. In any event, nothing of substance turns on whether theSITP is viewed as an improved version of the regular tax or as an improved version of a stand-aloneAMT.

38. Eric Toder, Changes in Tax Preparation Methods, 1993-2003, 107 TAX NOTES 759, 759(2005).

39. Joseph J. Thorndike, The Great Noncrisis of the AMT, 107 TAx NOTEs 245, 245 (2005).

40. Id.

41. Id.

1142 [Vol. 105:1133

The Theory and Practice of Tax Reform

ever. It certainly offends "an academic desire for tidiness, 42 to have two taxsystems when one should suffice-especially when the secondary tax sys-tem has strayed so far from its original purpose. And the Panel correctlynotes that some of the current features of the AMT are objectionable-

43especially the absence of inflation adjustments. If those features were re-moved, however, would there be any reason beyond aesthetics for makingAMT repeal a high priority? Actually, I think there would. Letting H&RBlock or TurboTax do the math may suffice for return preparation, but forplanning purposes taxpayers need to understand how they are affected bythe complex interplay of the AMT and the regular tax. Is one's marginal taxrate the regular tax rate or the AMT rate? If one makes an expenditure eligi-ble for a deduction or credit under the regular tax but not under the AMT,will one be able to claim the tax benefit? If taxpayers-unaware that theyare subject to the AMT-are induced to engage in behavior which is tax-favored under the regular tax but not under the AMT, the result is simplyunfair.4 The impact of the AMT on tax planning, not its impact on returnpreparation, is the best reason for repealing the tax, despite the Report'sfailure to mention that concern.

Having decided to propose repeal of the AMT, the Panel was faced withthe loss of $1.2 trillion of tax revenue within the ten-year budget window (p.43). That meant that some treasured tax breaks would have to be sacrificedto achieve revenue-neutrality. The president's instructions to the Panel sug-gested that the tax benefits for charitable contributions and forhomeownership were not to be eliminated, and the Panel wanted to increase,not decrease, opportunities for tax-favored savings. The deduction for stateand local taxes, therefore, was doomed. It was the only tax break that didnot seem to be off-limits, the repeal of which would raise enough revenue tocover most of the cost of the AMT repeal. Analyses performed by the Staffof the Joint Committee on Taxation suggested that repeal of the deductioncould cover close to two-thirds of the revenue cost of repealing the AMT, at

45least over the next few years.

42. Comm'r v. Duberstein, 363 U.S. 278, 290 (1960).

43. See supra note 37.

44. This is probably happening now, with respect to the tax credits for purchases of hybridautomobiles and certain other energy-efficient products. See Lawrence Zelenak, Of Prius Buyers,Blue States, Consumer Energy Credits, and the Alternative Minimum Tax, 109 TAX NOTES 657,659-62 (2005). Inducing desired behavior through the use of mirage-like tax benefits may be a cost-effective legislative approach (at least in the short run), but the cost effectiveness cannot justify theunfairness.

45. The Staff estimated that the AMT would raise $57.0 billion of revenue in 2006, $66.0billion in 2007, $81.4 billion in 2008, and $91.8 billion in 2009, for a total of $296.2 billion over thefour-year period. STAFF OF J. COMM. ON TAX'N, 109TH CONG., PRESENT LAW AND BACKGROUNDRELATING TO THE INDIVIDUAL ALTERNATIVE MINIMUM TAX II tbl. I (Comm. Print 2005). In its taxexpenditure analysis released a few months earlier, the Staff had estimated the reductions in tax

1143April 20071

Michigan Law Review

The Report advocates the repeal of the state and local tax deduction forthe same reason Willie Sutton robbed banks-because that's where themoney is. Perhaps viewing that explanation as insufficiently high-minded,however, the Panel offers a different justification in the Report:

This deduction provides a federal tax subsidy for public services providedby state and local governments. Taxpayers who claim the ... deductionpay for these services with tax-free dollars. These services ... represent asubstantial personal benefit to the state or local residents who receive them.... The Panel concluded that these expenditures should be treated likeany other nondeductible personal expense, such as food or clothing, andthat the cost of those services should be borne by those who want them-not by every taxpayer in the country. (p. 83)

This is too glib. Certainly some taxes can be viewed as payments forgovernment services received, and as to those taxes the Panel's argument issound. But some state and local taxes are redistributive-think of taxes usedto finance state expenditures under TANF (Temporary Assistance for NeedyFamilies) or Medicaid, and property taxes used to finance public schoolsthat the taxpayer's family does not use; as to those, the Panel's argumentdoes not apply. The inadequacy of the argument is so obvious it is surprisingthat the Report does not make a more serious attempt to defend its pro-posal-perhaps by offering evidence that the bulk of state and local taxesare taxes-for-services rather than taxes-for-redistribution, or by suggestingsome rough-justice rule for distinguishing nondeductible taxes of the formersort from deductible taxes of the latter sort. A long-standing deduction witha considerable constituency (consisting of both taxpayers claiming the de-duction and state and local governments politically able to impose highertaxes because of the deduction), and a plausible policy justification (at leastin part) deserves more respectful consideration before it is repealed.

The explanation for the inadequate explanation may be that the Panelthought the state and local tax deduction was doomed to a slow death undercurrent law, as a result of the nondeductibility of state and local taxes underthe ever-growing AMT. If taxpayers despair for the deduction even withouttax reform, they may not require a very convincing rationale for hasteningits demise. The Report suggests as much when it notes that "the AMT isincreasingly erasing the benefit of the state and local tax deduction for manymiddle-class taxpayers" (p. 84).

liability caused by the deduction of nonbusiness state and local taxes (including property taxes onowner-occupied homes) at $51.8 billion in 2006, $47.3 billion in 2007, $46.7 billion in 2008, and$48.4 billion in 2009, for a total of $194.2 billion. STAFF OF J. COMM. ON TAX'N, 109TH CONG.,ESTIMATES OF FEDERAL TAX EXPENDITURES FOR FISCAL YEARS 2005-2009, at 33 tbl.1, 39 tbl.1(Comm. Print 2005) (providing separate estimates for property taxes on owner-occupied homes andfor all other nonbusiness state and local taxes).

1144 [Vol. 105:1133

The Theory and Practice of Tax Reform

D. Home Mortgage Interest and Health Insurance:

Tilting at Two Windmills

Two of the Panel's recommendations are especially admirable on themerits, and especially unlikely to be adopted by Congress: the proposal toreplace the home-mortgage interest deduction with a 15% credit for a lim-ited amount of home-mortgage interest (pp. 70-75), and the proposal tolimit the exclusion for employer-provided health insurance to the averagecost of health insurance while allowing a similarly capped deduction forindividually purchased health insurance (pp. 78-82).

The reduction of the tax benefit for home-mortgage interest would besevere for many taxpayers-a surprising result, given the president's re-quirement that the Panel's proposals "recogniz[e] the importance ofhomeownership ... in American society.' 46 Consider, as a fairly typical ex-ample, a taxpayer with a $400,000 mortgage at an interest rate of 5% and amarginal tax rate (under current law) of 28%, living in an area where theaverage cost of a home is $300,000. Under current law, the taxpayer canclaim his entire $20,000 interest expense as an itemized deduction, produc-ing a tax savings of $5,600. Under the Panel's proposal, only $15,000 of hisinterest expense would be eligible for the credit (i.e., the interest on theprincipal amount of the loan, capped by the $300,000 average price of areahomes), and the amount of the credit would be only $2,250 (15% of$15,000). Other taxpayers would experience results much worse than this. Ataxpayer with a $1,000,000 mortgage with an interest rate of 5%, subject toa marginal tax rate of 35% and living in an area with an average home priceof $230,000, would go from a tax savings of $17,500 under current law(35% of $50,000 of deductible interest) to a credit of only $1,725 (15% ofthe interest of $11,500 on principal of $230,000). If the language in the Ex-ecutive Order was intended as code for "leave the home-mortgage interestdeduction alone," the president should have been more explicit, as the Panelfailed to crack the code.

The Report makes a strong case for its proposal on the merits, beginningwith the suggestion that "the tax code [may] encourage[] overinvestment inhousing at the expense of other productive uses" (p. 70). The Report goes onto note that the current deduction "exceed[s] what is necessary to encouragehome ownership or help more Americans buy their first home," and that "the$1 million mortgage limit may encourage taxpayers to purchase luxury resi-dences and vacation homes" (p. 71). It also notes the uncomfortable (todefenders of the current deduction) facts that 55% of the benefit of the de-duction is enjoyed by the 12% of taxpayers with cash incomes of $100,000or more (p. 72), that 46% of taxpayers paying home-mortgage interest ob-tain no tax benefit because they claim the standard deduction (p. 74), andthat a number of countries without tax benefits for home-mortgage interesthave homeownership rates comparable to that of the United States (p. 72).

46. Exec. Order No. 13,369, 70 Fed. Reg. 2323 (Jan. 7, 2005).

April 20071 1145

Michigan Law Review

Based on all this, the Report reasonably claims that its proposed credit,unlike the current deduction, "would encourage home ownership, not bighomes" (p. 73).

But the merits will not matter, as this particular proposal was evendeader on arrival than the Report as a whole.47 The National Association ofRealtors estimated that home prices would fall by 15% if the proposal wereenacted,48 the Mortgage Bankers Association labeled the proposal "a taxincrease for a lot of working Americans, 49 and the National Association ofHome Builders released survey results indicating that 75% of likely votersopposed the proposal.f Neither the president nor Congress has indicated anyinterest in the proposal. The only hope for a more rational homeownershiptax policy may be sustained inflation-the $1 million statutory limit on "ac-quisition indebtedness" generating deductible interest expense is notadjusted for inflation,5 and Congress has shown no inclination to increasethe ceiling in response to its erosion by inflation.

The story of the health insurance recommendation is basically thesame-another meritorious but doomed proposal. In a tax system designedto exclude from the tax base the costs of subsistence, the proper treatment ofhealth insurance is clear-the cost of basic health insurance coverage shouldbe excluded from the tax base, regardless of whether one obtains the cover-age through employment or purchases the coverage on the individualmarket, and there should be no tax benefits for the additional costs of"Cadillac" health insurance.52 By limiting the exclusion for employer-provided health insurance to the average cost of health insurance, and byproviding a similarly capped deduction for the cost of individually pur-chased health insurance,53 the Report gets this exactly right. Naturally, theReport's health insurance proposals are quixotic. America's Health Insur-ance Plans (AHIP) quickly released poll results showing that over 90% of

47. In the words of the Munchkin coroner, the proposal was "not only merely dead," but"really most sincerely dead."

48. Eduardo Porter & David Leonhardt, Tax Plan Could Raise Homeowners' Costs, N.Y.

TIMEs, Nov. 3, 2005, at C1.

49. Id.

50. Sandra Fleishman, Tax Proposal May Not Float, but It Sure Is Making Waves, WASH.

POST, Nov. 12, 2005, at Fl.

51. I.R.C. § 163(h)(3)(B)(ii) (2000 & Supp. 2006).

52. For a defense of these propositions, see Lawrence Zelenak, Of Head Taxes, IncomeTaxes, and Distributive Justice in American Health Care, 69 LAw & CONTEMP. PROBS. 103, 109-12(2006).

53. Under current law, individually purchased health insurance can be deducted as a medicalexpense under I.R.C. § 213, but only to the extent total medical expenses exceed 7.5% of the tax-payer's adjusted gross income, and even then only if the taxpayer itemizes deductions rather thanclaims the standard deduction. I.R.C. § 213(a).

1146 [Vol. 105:1133

The Theory and Practice of Tax Reform

voters opposed capping the exclusion for employer-provided insurance,54

and President Bush specifically rejected the proposal a few months later.55

In his 2007 State of the Union address President Bush proposed-surprisingly-a reform of the tax treatment of health insurance similar to the

56Panel's proposed reform. The immediate negative reaction from leadingmembers of the Democratically-controlled Congress suggests the proposalhas little or no chance of becoming law.57

III. THE TAX REFORM BIG PICTURE

A. Not with a Bang, but with a Whimper

The Panel's Report focuses attention on a major tax reform debate. Overthe past several decades, the core of the tax reform debate has been betweenproponents of taxing income and proponents of taxing consumption. Thereis almost universal agreement that the federal tax system should be at least1 8

modestly progressive, and no one (except Congress59 ) is in favor of a loop-hole-ridden tax base. There has been tremendous disagreement, however,over whether the tax base should move closer to the Haig-Simons incometax ideal,6

0 or whether the attempt to tax income should be abandoned in• - 61

favor of consumption taxation.Given the status of income-versus-consumption as the issue in the tax

reform debate, it is remarkable that the Panel presents both an option whichit labels an income tax and an option which it carefully refrains from callingan income tax, while expressing no preference between them. Moreover,each proposal is really an income-consumption tax hybrid. The SITP is even

54. Voters Polled in Bellweather [sic] States Shows Majority Reject Healthcare Tax Propos-als, MANAGED CARE Bus. WK., Nov. 15, 2005, at 39.

55. R.J. Lehmann, Health Insurance Advocates Applaud Rejection of Tax Plan Plank,BESTWIRE, Jan. 17, 2006.

56. 'The State of Our Union Is Strong; Our Cause in the World Is Right,' N.Y. TIMES, Jan.24, 2007, atA16 (extensive excerpts from the State of the Union address).

57. Robin Toner & Robert Pear, Bush Revives Some Past Proposals and Offers a New Initia-tive on Health Insurance, N.Y. TIMES, Jan. 24, 2007, atA15.

58. Even proponents of single-rate tax systems tout the average rate progressivity producedby exemptions or rebates built into their systems. See, e.g., NEAL BOORTZ & JOHN LINDER, THEFAIRTAX BOOK 81-90 (2005); ROBERT E. HALL & ALVIN RABUSHKA, THE FLAT TAX 124 (2d ed.1995).

59. See infra text accompanying note 65.

60. Haig-Simons income is "the algebraic sum of (1) the market value of rights exercised inconsumption and (2) the change in the value of the store of property rights between the beginningand end of the period in question." HENRY C. SIMONS, PERSONAL INCOME TAXATION 50 (1938).

61. The debate is too voluminous to cite more than a few examples here. In the legal litera-ture, significant contributions include William D. Andrews, A Consumption-Type or Cash FlowPersonal Income Tax, 87 HARV. L. REV. 1113 (1974); Barbara H. Fried, Fairness and the Consump-tion Tax, 44 STAN. L. REV. 961 (1992); Edward J. McCaffery, The Uneasy Case for Wealth TransferTaxation, 104 YALE L.J. 283 (1994); and Alvin C. Warren, Jr., Fairness and a Consumption-Type orCash Flow Personal Income Tax, 88 HARV. L. REV. 931 (1975).

1147April 2007]

Michigan Law Review

less of an income tax (and thus more of a consumption tax) than current law,and the GITP has enough vestigial income tax features that the Panel, hav-ing refrained from calling it an income tax, also refrains from calling it aconsumption tax.

Regardless of the fate of the Panel's proposals in the short term, I sus-pect that the proposals-especially the SITP-foreshadow the outcome ofthe great tax base debate. Official abandonment of income taxation isprobably politically impossible, both because of the difficulty of explainingto the electorate why those who work hard for their wages should be taxedwhile those who sit back and collect investment income should not, and be-cause of the massive transition problems in switching to a fundamentallydifferent system. On the other hand, much can be done to move the incometax closer to a consumption tax without the public relations or transitionproblems inherent in the switch to a pure consumption tax. The tremendousexpansion of Roth IRA-type savings vehicles proposed in the SITP is per-haps the most politically promising strategy for slouching toward aconsumption tax (without ever getting all the way there). This was recog-nized by the Bush administration at least as early as 2003, when it firstproposed the "Retirement Savings Accounts" and "Lifetime Savings Ac-

62counts" on which the Panel based its SFR and SFF accounts. Although theadministration's ongoing push for these accounts has not yet succeeded, theaccounts are politically plausible in a way that "pull[ing] the income tax outby its roots and throw[ing] it away" is not.63 This may be the way the in-come-versus-consumption debate ends-not with a bang, but with awhimper.

B. Tax Reform Can Succeed, but Only Because It Will Fail

To those who have followed United States income tax policy for severaldecades, the three desiderata of the Report's title-"Simple, Fair, and Pro-Growth"-are very familiar. In 1984, the Treasury Department issued a re-port which played an important role in the development of the Tax ReformAct of 1986. The Report was titled Tax Reform for Fairness, Simplicity, andEconomic Growth.64 The 1984 and 2005 Reports are similar in substance aswell as in name (although the 1984 Report is considerably more committedto the income tax ideal than the 2005 Report). Whenever experts issue a dis-passionate report on income tax reform, they always favor making the

62. Patti Mohr, White House Begins Selling Its Tax Cut to Congress, 98 TAX NOTES 631,633-34 (2003). The administration proposes these accounts annually. For the current proposal, seeU.S. DEP'T OF THE TREASURY, GENERAL EXPLANATIONS OF THE ADMINISTRATION'S FISCAL YEAR

2007 REVENUE PROPOSALS 5-10 (2006).

63. John Godfrey, Archer Keen on Killing Code; Full Speed Ahead on Tax Reform, 70 TAX

NOTES 1431, 1431 (1996) (quoting remarks of Ways and Means Chairman Bill Archer at a pressconference on March 4, 1996).

64. U.S. DEP'T OF THE TREASURY, TAX REFORM FOR FAIRNESS, SIMPLICITY, AND ECONOMIC

GROWTH (1984).

1148 [Vol. 105:1133

The Theory and Practice of Tax Reform

system simpler, fairer, and more pro-growth than current law, and they al-ways conclude that a broader-based, lower-rate tax is the way to achievethose goals. To state the obvious, the obstacles to making the tax systemsimpler, fairer, and more efficient were not technical in 1984, and they arenot technical today. The problems are political. After Congress succeeded inenacting a relatively simple, fair, and efficient income tax in 1986, itpromptly set to work dismantling its accomplishment. In his remarks to thePresident's Panel, Milton Friedman offered a cynical-but very plausible-view of how the 1986 Act came about, why it unraveled, and why the timemight soon be ripe for legislation resembling the 1986 Act:

The function of taxes is as a way in which ... legislators can raise cam-paign funds and that's widely recognized. A seat on the Ways and MeansCommittee is a very valued seat because it's a good way to raise funds....[A]s of 1986... the tax system had gotten so complicated, you had filledup the blackboard essentially so that Congressmen had nothing more tosell, and they were therefore willing to wipe the slate clean and start overagain.... And maybe it's time again for another wiping the slate clean andgetting rid of. . . our tax system [which] ... everybody agrees ... is ob-

65

scene.

Whether advocates of a fairer, simpler, and more efficient tax systemshould be cheered or disheartened by Friedman's analysis is debatable.The good news is that real tax reform will be possible when the Codehas become so loophole-ridden that the tax-writing committees have lit-tle left to sell to campaign donors; and the Code may soon attain thatcondition, thus making the Panel's Report particularly relevant for thenear future. The bad news, of course, is that Congress will broaden thebase only so that it can begin again to narrow it; there is no hope that theaccomplishments of tax reform will endure.

65. Sixth Meeting of the President's Advisory Panel on Federal Tax Reform 117-18 (2005)(remarks of Milton Friedman, Senior Research Fellow, Hoover Institute, Stanford University).

1149April 2007]

1150 Michigan Law Review [Vol. 105:1133